3rd Harmonics et al 3

[cjburns]

I am looking for a lead/link or information so as to expand my very thin knowledge of induced Harmonics in Computer Intensive electrical distribution systems in commercial premises.
Any Info would be of assistance as I knowledge challenged in this area.
CJ

 

[electricpete]

I don't have any commerical power experience, but I believe I understand the basic issue. Forgive me if I am telling you what you already know.

Electronic devices draw non-sinusoidal current which is periodic at 60hz. It will contain some harmonics (2nd, 3rd etc).

If we hook up identical electronic devices on each of three phases, then the currents drawn in each of the three phases will be identical except that they are shifted in time by one third of an ac cycle =(1/3)*(1/60)sec.

This means the first harmonics will be separated from each other by (1/3)*(1/60)sec times 360 degrees/(1/60sec) = 120 degrees.

Second harmoics will be separated by (1/3)*(1/60)sec times 360 degrees/(1/2*60sec) = 240 degrees.

Third harmonics will be separated by (1/3)*(1/60)sec times 360 degrees/(1/3*60sec) = 360 degrees.... ie the third harmonic currents from the three individual phases are all in phase with each other. When they meet at the neutral they will not cancel out (like 1st harmonic), but they will add together algebraically so the neutral 3rd harmonic current will be 3 times the phase third harmonic current.

Neutral wire must be sized to accomodate that current.

 

[jbartos]

Suggestion: Visit

http://www.powerquality.net/problem.htm

http://www.powerquality.net/solution.htm

http://www.ibus-phoenix.com/prod_power_conditioning.htm

http://www.powerquality.org/

etc. for more info

 

[electricpete]

I took a quick look at jbartos' link. I guess I forgot to mention that the harmonics affect all the supply equipment (but neutral wire is worst). So transformers, phase cables etc also need to be derated. How much? ... I can't say offhand.

 

[cjburns]

Thanks fgor all the tips gents/ladies?. Any clues as to screen flicker on PC screens?

 

[cjburns]

What is worse? Harmonic Voltage or Harmonic Current?

 

[electricpete]

A typical (and fairly realistic) way to model non-linear loads is to add a harmonic current source in parallel with the fundamental load (harmonic current source magnitude is a fixed complex multiple of the fundamental current.).

So the load determines it's own harmonic current. It must draw that current from the power system. This affects other loads only to the extent that the system voltage becomes distorted. The amount of voltage distortion is related to the product of the harmonic current magnitude and the source impedance. Since source impedance on per unit will typically be much less than one on per-unit basis, we expect the current distortion caused by a nonlinear load as measured at the load terminals to be much more than the resulting voltage distortion at those terminals (that wasn't intended to be a rigorous proof). In a current post on these forums it was reported that THD_V and THD_I at generator terminals were both approx 2%. At generator terminals we expect even more for voltage distortion to be less than current distortion (if distortion caused by load) since the source impedance is very low at generator terminals. In that case it seems reaonable to conclude that the generator is the source of the harmonics (not the load).

So my short answer would be:

voltage distortion is the key indicator as it relates to one non-linear piece of equipment effecting another load.

current distortion is the key indicator as it relates to overheating of source transformers, cables etc.

 

[jbartos]

Suggestion: It depends.
1. If you are concerned with power supplies, it is voltage harmonics.
2. If you are concerned with loads, it is current harmonics.
3. Current harmonics of a nonlinear load have a good potential to distort the power supply voltage and cause problem to other loads and irritate neighbors. In this sense, they are very bad and may be checked or monitored by the Utility.